Spatio-temporal patterns and climate-associated risk of elevated somatic cell count in dairy herd under tropical conditions.

Somatic cell count (SCC) is a widely used indicator of udder health and milk quality in dairy cattle; however, its spatio-temporal dynamics under tropical production conditions remain poorly understood. This study aimed to characterize spatio-temporal patterns of herd-level SCC and to evaluate associations between climatic conditions and SCC variation in Northern Thailand. An observational longitudinal study was conducted using 5,045 monthly herd-level observations from 102 dairy farms in Chiang Mai between January 2022 and August 2025. SCC was log10-transformed and analyzed as the primary outcome. Monthly meteorological data were obtained from the Thai Meteorological Department, and one-month lagged climatic variables were evaluated to capture potential delayed associations. Generalized additive mixed models were used to account for seasonality, long-term temporal trends, farm-level heterogeneity, and spatial structure using georeferenced farm locations. Inclusion of meteorological covariates improved overall model fit (ΔAIC ≈ - 11), although none of the individual climatic variables showed a statistically significant independent association with SCC at the 0.05 level. Addition of the spatial smooth further improved model fit (ΔAIC ≈ - 27 relative to the climate-adjusted model), and a significant broad spatial pattern persisted after adjustment for temporal, climatic, and farm-level factors. The final model explained approximately 52.5% of total deviance (adjusted R² = 0.51). In contrast, milk composition traits showed no significant residual spatial structure after adjustment. These findings demonstrate temporal and geographic heterogeneity in herd-level SCC under tropical dairy production conditions, while the associations of individual climatic predictors should be interpreted cautiously.